1.0 INTRODUCTION. 1.1 Prevalence of substance use. 1.2 Factors associated with substance use

Transcription

1 1.0 INTRODUCTION Alcohol misuse is one of the most significant public health concerns facing South Africa today. Globally, cannabis is the most widely used illicit drug, with an estimated 144 million people using it annually. It constitutes the main drug of abuse in Africa (2). Substance abuse is associated with employee illness, occupational accidents, increased health services utilisation, and decreased productivity (3,4). However, despite the fact that South Africa is one of the major mining countries in the world, local research on alcohol and cannabis use among mineworkers is limited. 1.1 Prevalence of substance use It is estimated that 6% to 16% of the average workforce is likely to be alcohol dependent and that a further 20% is likely to experience drug related problems (5,6). In South Africa, the prevalence of alcohol dependence among adults is estimated as 10%, while that of risky drinking among workforces such as the mining industry has been estimated at 25% or more (7). In a South African gold mine, the prevalence of risky drinking among workers was found to be 32% and the majority of these employees were in unskilled or semiskilled occupations (8). In another study carried out in South Africa, the highest rates of alcohol abuse as a household problem (32%) were reported among unskilled manual workers, while the lowest rates (9.1%) occurred among professionals (9). Among miners in Argentina, 34% were found to be weekly alcohol drinkers, while 65% chewed coca leaves daily (10). In 1984, the Addiction Research Foundation in Canada reported that 11% of adults in Ontario above 18 years old used cannabis (11). 1.2 Factors associated with substance use Historically, practices in the Mining and Agriculture industries such as the dop system, migrant labour system, availability of cheap or free alcohol, and availability of alcohol on credit, may have contributed towards increased alcohol use in the South African workforce. The dop system, officially prohibited in 1961, entails payment of workers with alcohol in lieu of wages (12,13). Factors, which may contribute to cannabis use, include the fact that it is inexpensive, easy to procure, prosecution is infrequently enforced, and is perceived by many not to be problematic. Poverty, boredom, and inadequate health education, have also been associated with substance use (14). In a South African gold mine the lifestyle of miners such as living apart from families for prolonged periods was found to encourage unhealthy alcohol consumption (14). Higher rates of alcohol use have been found among miners who have only ever worked underground compared to those who work aboveground, and among miners with a heavy workload (10). Daily use of coca was also found to be significantly higher among miners with a heavy workload (10). Stressful

2 working conditions as are found underground, and heavy workloads may encourage alcohol and drug use, which may serve as a coping mechanism (15,16). Stress, loneliness, and boredom have also been cited as reasons for alcohol use by South African mine workers (8). 1.3 Effects of substance use Effects of alcohol use Absenteeism, sick leave, and accidents have been found to be higher among workers who use excessive alcohol (3,17). In a South African pulp mill, blood alcohol was found to be positive in 18% of cases of injury, while in a copper mine in Zambia, blood alcohol was positive in 30% of accident cases (17,18). Excessive alcohol use is also associated with social problems like violence, and can predispose to illnesses such as hypertension, gastritis, liver cirrhosis, gout, tuberculosis, and physical dependence with withdrawal symptoms, and depression (19) Effects of cannabis use Regular cannabis use has been associated with impaired social and occupational functioning (20). The primary psychoactive constituent is delta-9- tetrahydrocannabinoid (THC) (21). Cannabis use results in feelings of euphoria and relaxation, and acute effects include impairment of attention and shortterm memory, and loss of coordination (22,23,24). Chronic effects include psychological dependence characterised by deterioration in psychosocial functioning; subtle cognitive deficits, particularly attention, learning, and executive functioning (organising and integrating of information); possible triggering of onset of schizophrenia; increased vulnerability to respiratory illnesses; impaired lung function; and precancerous changes in lung tissue (1,20). 1.4 Screening tools for substance use Screening tools for alcohol dependence The Diagnostic and Statistical Manual of the American Psychiatric Association, 4 th edition (DSM-IV), defines alcohol abuse as a pattern of use which leads to clinically significant impairment or distress, as manifested by one (or more) of the following in a 12-month period (7):? Recurrent alcohol use resulting in a failure to fulfil major role obligations at home, school, or work? Use of alcohol in situations in which it is physically hazardous (e.g. driving a car)? Recurrent alcohol use leading to legal problems (e.g. drunken driving)? Continued alcohol use despite persistent or recurrent social or interpersonal problems caused or exacerbated by alcohol. 2

3 Screening tools for alcohol misuse include the CAGE, the AUDIT (Alcohol Use Disorder Identification Test), and the brief MAST (Michigan Alcohol Screening Test) questionnaires (25,26,27). They are specific and reliable, and help to screen individuals who require further assessment for alcohol dependence. The brief MAST is an abbreviated version of the original 25-item MAST published by Selzer in 1971, and like the AUDIT, it is also a 10-item questionnaire (7). The CAGE questionnaire was developed by Ewing and Rouse in Comprised of the following four questions, it is easier to administer (28):? Have you ever felt you ought to Cut down on your drinking?? Have people Annoyed you by criticising your drinking?? Have you ever felt bad or Guilty about your drinking?? Have you ever had a drink first thing in the morning to steady your nerves and get rid of a hangover? (Eye-opener) Two or three positive responses are highly suggestive of alcohol abuse and possible dependence, while four positive responses are virtually diagnostic. Laboratory tests or test combinations that can be used for screening alcohol abuse include mean cell volume (MCV), aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma glutamyl transferase (GGT), and uric acid (28). Breathalyser testing can be carried out to detect acute intoxication but cannot assess chronic misuse. On ingestion, alcohol is rapidly absorbed from the upper gastrointestinal tract. Peak concentrations of ethanol are attained approximately one hour after ingestion and factors influencing levels attained include the rate at which the drink was taken, whether it was consumed with food, rate of gastric emptying, and body habitus (28). Between 2% to 10% is eliminated in urine and breath (28) Screening tools for cannabis use Marijuana is usually smoked but may ingested, either incorporated into food, or as a liquid extract (tea). It is rapidly absorbed from the lungs into the blood with quick onset of effects. When ingested however, onset is slower but effects more prolonged. The natural metabolites of cannabis (cannabinoids) are found in blood, bile, faeces, and urine. It may be detected in the latter within hours of exposure (30). These metabolites being fat soluble, are stored in the body s fatty tissues including the brain, for prolonged periods after use (30). It may be detected in urine months after last exposure, depending on the frequency and intensity of use (31). Qualitative screening for detecting cannabinoids in urine can be carried out using commercially available rapid tests and laboratory tests with varying levels of reported sensitivity and specificity. However, confirmatory laboratory tests, which also quantify the amount of cannabinoids in the urine, exist, of 3

4 which the preferred method is the Gas Chromatography/Mass Spectrometry (GC/MS) method (31). 1.5 Measures for control of substance use among mine workers The South African Mine Health and Safety Act of 1996 states that an employer must provide conditions for safe operation, and every employee must take reasonable care to protect their own health and safety, and that of other workers who may be affected by an act of omission on their part (32). It also states that no persons in a state of intoxication, or in a state likely to render him incapable of caring for himself or others in his charge, will not be allowed to enter a mine. This is also stipulated in Regulation of the Minerals Act 50 of 1991 (33). However, there are no clear guidelines for implementation and the level of interpretation of this responsibility varies from mine to mine; from those in which there are no clear substance use guidelines, to those with draft policies, and to those with policies. Where policies exist, they describe the mine regulations in terms of substance use, under what circumstances testing will be carried out, how it will be carried out, and how results will be dealt with. In New South Wales most mines have an alcohol policy, which may include random testing, pre-shift self-breathalyser testing, and awareness programs (34). Buy-in of stakeholders is however of utmost importance in any control program. In 1995, the International Labour Organisation (ILO) adopted a code of practice on the management of alcohol and drug related issues in the workplace (35). This code emphasises a preventive approach and embraces the following principles:? Joint assessment by employers, workers and their representatives of the effects of drug use on the workplace and their cooperation in developing a written policy for the workplace? Consideration of alcohol and drug related problems as health problems, and a need to deal with them without discrimination, like any other problem in the workplace? Recommendation that drug and alcohol policies should cover all aspects of prevention, reduction, and management of alcohol and drug related problems, and integration of relevant information, education and training programs where feasible, into broad-based human resources development, working conditions, or occupational safety and health programs. 4

5 ? Establishment of ethical principles which are vital to concerted and effective action, such as confidentiality of personal information, and the authority of the employer to discipline workers for employment-related misconduct, even where it is associated with the use of alcohol and drugs.? Consideration of fundamental legal, ethical, and moral issues involved in testing body fluids for alcohol and drugs and determination of when it is fair and appropriate to carry out such testing. The Occupational Alcohol Program (OAP) of the 1970s was one of the earliest attempts at addressing alcohol misuse in the workplace (36). This has been replaced in recent times by Employee Assistance Programs (EAPs) which are broader based and aim at addressing all personal problems that are affecting, or that have a potential to affect an employee (6,37). 1.6 Motivation for this study Evaluating the prevalence, knowledge, and practice of alcohol and cannabis use among mine workers in South Africa in relation to health and safety, will help to find out more about substance use among this population and assist in development of recommendations to improve health and safety. This is of importance to the mining industry as this ultimately impacts on productivity and finances. The cost of alcohol and drug abuse to South Africa has been estimated at R 2 Billion per year (38, 39). This study aims to provide evidence on which health intervention strategies can be based. 1.7 Study aim and objectives Overall Aim To determine the prevalence and factors which influence alcohol and cannabis use among mineworkers in South Africa Specific objectives? To determine the prevalence of alcohol and cannabis use.? To determine the knowledge, attitudes, and practice regarding alcohol and substance (cannabis) use amongst miners, and its relationship to health and safety.? To determine factors which influence alcohol and cannabis use.? To make the findings available to all stakeholders, so that appropriate recommendations can be implemented. 5

6 2.0 METHODS 2.1 Study design This is a cross-sectional analytic study. 2.2 Study site description Selection of study sites Eleven mines were purposively selected to represent the major commodities mined in South Africa, size and geographical distribution, and were classified into five commodity groups namely platinum, diamond, gold, coal, and other category comprising a granite mine. These mines comprised three platinum mines, three gold mines, two diamond mines, two collieries, and one granite mine. Of these eleven mines, four mines declined participation (one each in the gold, platinum, diamond, and coal category), bringing the total number of mines in which the study was carried out to seven i.e. two platinum mines, two gold mines, one diamond mine, and one colliery, and one granite mine. Although the mines that declined participation did not officially state reasons for doing so, their reasons may not have been unrelated to concerns about anonymity and job security of participants, raised during consultation with employee representatives (section ). The mines in the gold and platinum categories were not replaced as the number of consenting mines in these categories was deemed adequate. The mines in the other categories were not replaced due to logistics of finding replacement mines within the timeframe of this project. In order to ensure confidentiality of information obtained from this study, individual names of mines and their geographical locations are not described in this report and have been coded as follows: Platinum commodity Mine P1 (small size) Mine P2 (large size) Gold commodity Mine G1 (small size) Mine G2 (large size) Diamond commodity Mine D1 (small size) Coal commodity Mine C1 (small size) Other commodity (granite) Mine O1 (small size) 6

7 2.2.2 Background information on study mines Table 2.1 displays a summary of background information on study mines. Table 2.1: Summary of background information on study mines Background Mine information P1 P2 G1 G2 D1 C1 O1 When commissioned Underground (UG) or opencast (OC) OC UG OC UG OC UG OC Employee size Type of employee lodging All live in surrounding towns Hostels available All live in surrounding towns Hostels available All live in surrounding towns Hostels available All live in surrounding towns Substance use policy Draft policy Code of conduct Code of conduct Policy document Policy document Policy document Code of conduct Substance use policies of study mines While some mines have substance use policy documents, others have informal codes of conduct, which govern their practice concerning substance use among employees Mine P1 Draft policy guidelines at this mine govern substance use among employees. Possession or use of substances of abuse in the workplace is prohibited. The suggested alcohol limit for employees reporting for duty is 0.2 mg/1000ml of breath and employees under the influence of substances that may impair performance of normal duties are not permitted to work. Tests for alcohol and other drugs are carried out as part of pre-employment screening, following incidents at work involving fatal injuries and damage to property, on reasonable suspicion of intoxication, when employees are found in possession of substances, and following computerized random selection at the mine entrance. However, employees who suspect they are under the influence of substances may request voluntary testing before the start of their shift. 7

8 Disciplinary measures are meted out to employees who violate this code of conduct depending on the circumstances. Where an employee refuses to be tested, an inference of positive use of substances is made and such evidence can be used against such an employee during any disciplinary hearing that may follow. Employees with substance use problems can voluntarily inform management of such a problem (before it is discovered during any testing procedure) and enroll in an Employee Assistance Program (EAP), sponsored by the mine. Should there be a relapse, costs of a repeat rehabilitation program are borne by the mine, after which the employee assumes financial responsibility in cases of future relapse. Costs are also borne by the employee where they are found to be non-cooperative with rehabilitation or where they were caught out e.g. following random selection at the mine gate for drug tests Mine P2 Alcohol is not allowed on the mine premises except for special functions for which prior official permission has been obtained, and a mine official is present to take responsibility. A code of conduct allows for testing of employees who are suspected to be under the influence of substances. Where the result is positive, appropriate disciplinary action is instituted. The legal breath alcohol driving limit for non-professional drivers of 0.24mg/1000ml of breath in section 65 (5b) of the National Road Traffic Act 93 of 1996 is regarded as the limit for positive breathalyser results for alcohol (40). There is no system in place for pre-employment or random testing of employees. However, testing is carried out following accidents at work. Employees who are found to have chronic substance misuse problems may be considered for rehabilitation programs Mine G1 There is a code of conduct operational at this mine whereby employees who are suspected to be under the influence of substances are taken to the security department where tests are carried out. Should result be positive, a hearing is held and disciplinary procedures instituted. There is no random testing for substances, however testing of involved employees is carried out following accidents at the mine. The mine alcohol limit of 0.24 mg/1000ml (i.e. 0.05% Blood Alcohol Concentration) is the same as the legal driving limit of non-professional drivers Mine G2 A policy existent in this mine since August 2000 addresses substance abuse among drivers and heavy machinery operators. It reiterates the mine s commitment to ensuring a safe working environment and the responsibility of employees to their fellow workers in achieving this aim. Based on the Road 8

9 Traffic Act of 1996, the Mine Health & Safety Act of 1996, the Occupational Health & Safety Act of 1993, and the Labour Relations Act of 1995, this policy encompasses routine screening for chronic alcohol and cannabis abuse (and other drugs of abuse where necessary) in drivers and heavy machinery operators. Education of employees in this job category about substance abuse is incorporated into initial induction. Screening is carried out during preemployment and periodical medical testing, but random testing may be carried out if required. Consent is obtained before testing is carried out, and refusal to consent may result in inability to determine an employee s fitness for the job. The limit for positive breathalyser tests is 0.24 mg/1000ml. Drivers found to be misusing alcohol or drugs may be declared temporarily or permanently unfit to carry out their duties. EAPs exist for rehabilitation and the costs of a first time rehabilitation are borne by the medical aid. Once cleared by EAP, the employee is retested before being declared fit to return to work. Thereafter, testing is undertaken at random intervals. If there is any relapse, the employee is declared permanently unfit to drive. Testing for acute alcohol intoxication whilst on duty is not covered by this policy but by a disciplinary code of practice Mine D1 A substance use policy has been existent at this mine since August It states that according to the Minerals Act, no person in a state of intoxication, or in a state likely to render him incapable of caring for himself or others in his charge, will be allowed to enter a mine. Employee testing is carried out following suspicion of being under the influence of substances at work, on discovery of substances in an employee s possession, and after involvement in work related accidents. A breathalyser result of 0.24 mg/1000ml of breath is regarded as the limit for breath alcohol results at this mine. Though the right of an individual to refuse testing is acknowledged, a negative inference may be drawn in such an instance. Evidence obtained from testing may lead to disciplinary action. Financial assistance for a rehabilitation program at a rehabilitation centre is provided for employees who request treatment of medically documented dependence on substances, and for employees identified to have such problems. This assistance, which is for one course of treatment, is only repeated in exceptional circumstances Mine C1 The mine policy states that employees and visitors must not enter the workplace in a state of intoxication. Voluntary as well as random breathalyser tests are carried out to assess alcohol levels. Any reading higher than 0.00 mg/l is considered positive and sets a disciplinary procedure in motion. Employees with positive results are not allowed to work on the day of the test unless a second test performed an hour later, is negative. Should the second 9

10 test also be positive, the employee is asked to leave the premises and faces disciplinary action consequent to being absent from work without permission Mine O1 In terms of the Mines Health and Safety Act (1993) the mine reserves the right to do the following:? Conduct random breathalyser tests on any of its employees during working hours, at times and places and in a manner as decided by management to ensure compliance by the employees to the above requirements.? Have the necessary tests conducted on any employee suspected of being under the influence of an intoxicating substance thus ensuring that such individual is not wrongly accused. Any positive breathalyser result irrespective of the level of alcohol is regarded as positive by the mine. Where an individual refuses to be tested under these conditions it is assumed that the individual is in a state of intoxication and necessary disciplinary action is taken. However, there is at present no routine in place at this mine to carry out the above. 2.3 Sampling Sample size calculation The sample size for each of the five commodity categories (i.e. platinum, gold, diamond, coal, and other) was calculated as 385 by assuming 50% prevalence of drug use, 95% precision, and 5% margin of error (Appendix 1). An additional 20% of 385 (i.e. 77) was added to this sample size in case of refusals from the study, bringing the sample size per commodity to 462. However as there was only one mine in the other category, half of this sample size (i.e. 231) was allotted to it, bringing the total sample size of all the mines in the five commodities to 2079 [i.e. (462 X 4) + 231]. The sample size of 462 was proportionally distributed between the different mines in the four commodity groups (platinum, gold, diamond, and coal) with respect to the employee size in each mine at the onset of the study (January 2002). In each mine, the sample size was then proportionally distributed between contract workers and fulltime workers. In mines P1, G1, and D1, the sample size was adjusted, to ensure that the number of contract workers would not be less than 30 per mine so as to have adequate sample size for meaningful analysis of data. 10

11 Though one platinum mine (P3), one diamond mine (D2), and one coal mine (C2) eventually declined to participate, they had been included in the sample size calculation for their commodity group at the commencement of data collection. Due to lengthy negotiation process to obtain buy-in of stakeholders (section ), data collection was commenced at some consenting mines, while negotiations continued at other mines in the same commodity category, some of which later declined participation, making it too late to increase the sample size of other mines in the same category. The third gold mine (G3) was not included in the calculation because there was an early indication that buyin of stakeholders would not be obtained within the timeframe of this project. However, with the exception of coal commodity, the sample size of mines in other commodity groups were not unduly affected as adjustments had been earlier made to increase their estimated sample sizes due to smaller numbers of contract workers in comparison to full-time employees in these mines. Table 2.2 shows the sample sizes of study mines. Table 2.2: Summary of sample size of study mines Mine Contract workers Fulltime workers Total P P G G D C O1 Nil Total Selection of subjects A systematic sampling method was utilised. A register of all mine employees including management staff was obtained from each mine authority. Contract workers were grouped together on a separate list from fulltime workers. Where possible, employees were grouped together according to job category and workstation. Every nth employee was selected. This n th factor was determined by dividing the total employee size of each mine by the estimated sample size. However, due to withdrawal of certain sections of the mine from the study in mines P1 and O1, the n th factor was determined using an employee size of 716 and 400 respectively instead of 1500 for mine P1 and 656 for mine O1, as described in table 2.1 above. 11

12 2.3.3 Inclusion / exclusion criteria and replacement technique The four mines that declined to participate were excluded from the study. The eleven participants of a pilot study at mine P1 (section 2.5) were excluded from the main study that was carried out at a later date at this mine. Participants in sections of mine P1 and mine O1 who declined to participate were also excluded. All other employees were eligible to participate in the study including management staff and contract workers. Employees selected by systematic sampling were included in the study. Where an employee was unwilling to participate, the next name immediately below this employee on the employee register was selected. 2.4 Instruments of Measurement These include questionnaire, urine testing for cannabis, and breathalyser testing for alcohol Questionnaire Face-to-face structured interviews were carried out using a questionnaire (Appendix 5), administered by a trained team of interviewers who speak local languages. This questionnaire helped to determine the prevalence of alcohol and cannabis use among miners as well as their knowledge and practices with regard to alcohol and substance use, and their perception of its health and safety risks. It was structured to eliminate biases as far as possible. Quality and consistency were achieved through keeping the questions mainly closed and simple Breathalyser testing for alcohol Blood alcohol was assessed with the Alcatest 7410 plus RS breathalyser, the calibration of which has been verified against another breathalyser (Alcatest 7110), certified by the Council for Scientific and Industrial Research (CSIR), and the South African Bureau of Standards (SABS). This device uses disposable mouthpieces with one-way valves, such that air breathed into the instrument can only flow into the breathalyser, and cannot be inhaled by the participants, thus preventing transmission of infection. To avoid legal or ethical implications, where interviewers might be faced with the dilemma of allowing miners with excess breath alcohol levels to commence their work-shifts, the reading on the breathalyser screen was permanently set on a pass mode (i.e. normal blood alcohol level). Breath alcohol levels were automatically stored in the instrument, and actual results were downloaded onto a computer and read off after the interviews Urine cannabis testing Initial testing for Tetra-Hydro-Cannabiniod (THC), a metabolite of cannabis, was carried out using a THC test kit on-site. Further tests were carried out on randomly selected samples using the COBAS integra laboratory method. 12

13 Where there were discrepancies between results of the test-kit and the COBAS method, further tests were carried out using the gas chromatography method, which is the gold standard for THC testing. Having determined the sensitivity of the test kit to be 80% and specificity to be 97% in this manner, the use of this test kit was discontinued and further samples were tested at the laboratory using the COBAS method, which had a higher accuracy. 2.5 Pilot study A pilot study was carried out in mine P1 among 11 employees to pre-test the questionnaire and other data collection tools. The questionnaire was then modified as necessary. 2.6 Data collection Consent to carry out the research was obtained collectively from employee and management representatives in each mine through a consultative process with the research team (Appendix 2A) and from individual participants by trained research assistants (Appendix 2B). The purpose of the study was carefully explained to participants and written informed consent (Appendix 4) was obtained before interviews commenced. Anonymity was preserved, by excluding participants names on questionnaires, and keeping information obtained confidential. Participants were reassured that specimens obtained would only be tested for alcohol and cannabis. To facilitate the data collection process, trained research assistants were each given a pamphlet highlighting salient points in the process (Appendix 2B). The average data collection period per mine was 5 days Accessing of employees At all mines, selected employees were not aware beforehand that they would be asked to participate in the research and were only informed at the time of the test. Mine management representatives and in some cases Union officials, were notified about the day of our arrival so as to facilitate logistics for the data collection process. In order to facilitate accessing of employees, it was requested that employee lists classify workers according to workstation and shifts. Information about the number of shifts, time of commencement of shift and other information that would facilitate the data collection process was obtained for each mine (Appendix 3), and data collection spanned all shifts in study mines. Shifts range from one to four per 24-hour cycle depending on the mine, with some morning shifts commencing around 3.45am and some night shifts commencing around 10pm. While some employees are employed as part of a shift that works permanently at night or permanently during the day, some shifts rotate between morning, afternoon, and night duty. A list, classifying workers according to the shift facilitated accessing of workers who, though they may work at different times 13

14 of the day, usually rotate with their shift group. At some mines, aside from the shifts working at any point in time, there was an additional shift that was currently on a rest cycle of a few days to one week. To facilitate accessing of workers in all shifts during the data collection period, information was obtained about the two consecutive days that all shifts could be accessed at work (i.e. the days when the current rest shift is resuming duty and the new rest shift is going on break) Accessing of workers at mines with electronic access gates At surface and underground mines where electronic access gates were available, workers were accessed by parade technique. This implies that prior to the commencement of the shift, selected employees names were entered into the company computer system. On clocking in for duty, such employees were not allowed access into the mine and were requested by the mine Human Resource Officers controlling access into the mine to meet the research team waiting nearby, where they were invited to participate Accessing of workers at mines without electronic access gates At surface mines where no electronic access gates were available, employees were accessed shortly before their shifts at their workstations Timing of data collection The data collection process consisted of three parts (i.e. administration of a questionnaire and obtaining of breath and urine samples), which were to be carried out pre-shift without disruption of mine productivity. It was however thought that should it be impossible to carry out all three parts pre-shift due to time constraints, data collection could be split such that breathalyzer testing, the most crucial as alcohol could be metabolized in the body within hours (cannabis may still be detectable up to a month after use), would be done preshift while employees would be requested to report post-shift for structured interviews and urine samples Timing of data collection at surface mines At mines P1, O1, G1, even though the plan was to carry out the study pre-shift, it was sometimes necessary to continue data collection into the shift. This was due to time constraints in completing the process pre-shift as some employees arrived at work shortly before their shift, with concerns arising from the supervisor about interviewing several people at the same time and possible lateness of employees for pre-shift briefing and affectation of productivity. Requesting participants to arrive early on the day of the interview could lead to bias in the results of the breathalyser testing with possible modification of alcohol consumption the night before the interview. Hence, in instances where there was inadequate time to complete interviews before commencement of the shift, interviews started pre-shift and were sometimes staggered into the shift (i.e. when the persons being interviewed returned to work on completion of their interviews to relieve other employees, the next set of selected 14

15 participants were requested by the shift supervisor to come for interviews). However, delay of commencement of shift after blasting at the shaft sometimes facilitated pre-shift completion of data collection, when this occurred during the data collection period. At mine D1 where the available pre-shift period was only adequate to carry out breathalyzer testing, breathalyzer testing was done pre-shift by parading selected participants, while structured interviews and urine sampling were carried out during the shift Timing of data collection at underground mines Timing of data collection was more crucial at underground mines where employees are transported underground in an enclosure that runs on a strict schedule. The culture was different from shaft to shaft. At some shafts, the majority of workers arrived early and spent time chatting with fellow workers while awaiting commencement of their shift. However at other shafts employees arrived at work shortly before they were due to go underground leaving inadequate time to explain the study to participants in order to obtain breath samples pre-shift, while postponing structured interviews and urine sampling post-shift. At mines C1 and G2, all three parts of data collection were carried out at the shaft pre-shift as employees arrived early enough to complete the process. However, in mine P2, 15% of all respondents were seen post-shift due to logistic difficulties in accessing employees pre-shift. The alternative plan of carrying out breathalyser testing pre-shift, and structured interviews and urine testing post-shift could not be carried out as there was inadequate time to explain the study and obtain consent before employees had to go underground to resume their shifts Challenges encountered Lengthy consultation process A lengthy consultation process led to delayed commencement of data collection due to concerns of employees about anonymity, fairness in random selection of participants without involvement of management, victimisation and job security, and perceptions that samples obtained may be tested for Human Immunodeficiency Virus (HIV). These issues were addressed during several meetings with stakeholders at participating mines Logistics of accessing employees Parading of employees at underground mines was not a foolproof method as some employees did not respond to the parade despite several attempts, and there was no other means of contacting them at the time. Such employees were subsequently replaced, though some of these replacements did not respond as well. 15

16 Unavailability of urine sample at time of interview Employees were each given a specimen bottle and asked to fill it with urine in a nearby sanitary facility at the end of the structured interviews. Though they had initially consented to providing urine samples, when it came time to give the sample, a few participants said they did not feel like passing urine at the time. While some returned later to give a urine sample, it was in the main difficult to obtain these samples as most participants who did not give a urine sample in the first instance said they still did not feel like passing urine after further visits to their workstations at different times during the data collection period. A handful of samples were obtained from participants that resembled clear tap water Facilitating factors Cooperation of stakeholders The natural hierarchy of employee representatives was acknowledged and early buy-in of Union officials was obtained at the national, regional, and mine levels respectively. Some of the Union officials at mine level joined the research team during the data collection period, while others made their Union offices available for interviews, giving the team credibility with participants. Assistance of management through introduction to relevant personnel and in some cases releasing of an employee known and trusted by workers to chaperone the team during the data collection period, facilitated the process Masking of results on breathalyser screen The fact that the breathalyser screen did not reflect results making interviewers and participants unaware of results at the time of the test allayed some of participants fears about confidentiality Non-invasive nature of requested tests Participants were relieved that tests were painless and that samples requested did not include blood and saliva as they felt these could possibly be tested for HIV without their knowledge Experience gained by research team from mine to mine The data collection process became more efficient from mine to mine with experience gained from each mine. 16

17 2.7 Quality assurance Breathalyser testing High repeatability of tests Repeatability of tests carried out using the Alkatest 7410 plus RS breathalyser is high as only 1cc of breath is analysed every time a test is carried out irrespective of the amount of air that is blown into the mouthpiece Calibration of breathalyser The breathalyser was regularly re-calibrated during the data collection period to ensure accuracy of results obtained Pre-test mini questionnaire A mini questionnaire (Appendix 6) was designed to find out if participants had recently used substances (such as mouth sprays and cough syrups containing alcohol) that could affect accuracy of breathalyser testing so that adequate waiting time could be observed before testing Urine testing Selection of urine testing method As discussed in section above, comparison was made between the accuracy of dipstick testing and laboratory COBAS testing for cannabis. The latter test with a higher accuracy was selected for testing of samples Collection of urine specimen Participants were instructed to collect the first part of their urine stream into the sample bottle (Appendix 2B), as this is the part that is most suitable for THC testing Storage of urine samples Onsite, urine samples were stored in cooler boxes in which the temperature was maintained below 4?C for no longer than 48 hours, after which they were stored frozen in a freezer while they awaited analysis. This was to ensure that should THC be present in any urine sample, it remained biochemically stable till detection at the laboratory. 17

18 2.8 Data analysis Data was analysed using Excel and SPSS. Responses were coded and descriptive statistics were carried out. Cross tabulations yielding p-values were performed with chi-squares. 2.9 Possible limitations Ensuring truthful responses from the study population was the greatest challenge. This was addressed by early buy-in of all stakeholders including employee representatives and assuring participants of precautions taken to ensure confidentiality. The questionnaire was drawn up in English and translated to local languages in the field and there might have been small differences in meaning. However, the use of well trained local researchers well able to translate the questions into the language of the participant, and able to understand the answers (with the nuances of the different languages) ensured that participants responses were accurately represented. Although breathalyser and urine testing may be established ways of screening alcohol and drug abuse, a negative breathalyser test does not rule out chronic alcohol use (as the breathalyser only measures short-term use), and a positive test for cannabis does not mean that the worker is impaired at the time the sample is taken as metabolites of cannabis may remain in the urine long after the drug s effects have subsided (31). This study was carried out as a pilot study to find out information about alcohol and cannabis use among mine workers in South Africa. Results may not be generalisable to all the mines in South Africa, but will provide valuable information about alcohol and cannabis use among this population. 18

19 3.0 RESULTS OF BREATHALYSER AND URINE TESTING 3.1 Response rate Response rate varied from mine to mine between 84% and 99% and the replacement rate was between 3% and 8%. An additional 20% had however been added to the sample size from the onset in case of refusals. Nonrespondents include refusals & replacements not found at their workstations after several attempts. Within mines, there was a variation in the proportion of questionnaires administered and breath and urine samples obtained. The lowest is the number of urine samples obtained because some participants were unable to produce urine at the time of the interview. Table 3.1 describes response rate by mine. Table 3.1: Response rate by mine Mine Estimated sample size Questionnaires administered Breath samples collected Urine samples collected P (98.2%) 108 (98.2%) 108 (98.2%) P (87.3%) 200 (87.3%) 199 (86.9%) G (97.4%) 111 (97.4%) 110 (96.5%) G (84.7%) 359 (84.7%) 358 (84.4%) D (99.4%) 318 (99.4%) 305 (95.3%) C (98.9%) 266 (98.5%) 262 (97%) O (90%) 204 (88.3%) 200 (86.6%) Total (92.5%) 1566 (92.2%) 1542 (90.8%) Breathalyser and urine results are presented in sections 3.2 and 3.3, respectively. Results describe findings of individual mines and mean finding for all mines. 3.2 Breathalyser results The results in sections to below, describe all samples that contained alcohol irrespective of the level. All other samples not described contained no alcohol whatsoever (i.e. 0.00mg/1000ml). In the absence of a legislated alcohol limit in the mining industry for mine workers on duty, the current South African legal driving limit of 0.10mg/1000ml of breath for 19

20 professional drivers as described in section 65 (6) of the National Road Traffic Act of 1996 (40) was used as a benchmark for determining levels at or above which impairment of judgement is expected (. This is because heavy and complex machinery is often used in mines. In results presented, breath samples containing alcohol below this driving limit have a suffix?, those with breath alcohol levels equal to the driving limit of 0.10mg/1000ml are assigned?, while those above the limit are assigned? Breathalyser results for mine P1 Of the 108 breath samples obtained at this mine, only one sample (0.9%) contained alcohol, and it was below the legal driving limit for professional drivers of 0.10mg/1000ml. This result is displayed in table 3.2. Table 3.2: Positive breathalyser result for mine P1 Day of sample collection Time of sample collection Wednesday 8.50am 0.04? Positive breath alcohol result (mg/1000ml) Breathalyser results for mine P2 Of the 200 breath samples obtained from this mine, 4 samples (2%) contained alcohol. Three of these samples (1.5%) contained alcohol at or above the 0.10mg/1000ml driving limit. One of the samples obtained contained alcohol at a level (0.50mg/1000ml) five times the 0.10mg/1000ml limit. Positive breath alcohol results are displayed in table 3.3. Table 3.3: Positive breathalyser results for mine P2 Day of sample collection Time of sample collection Saturday 1.53am 0.27? Monday 9.17pm 0.07? Tuesday 2.15pm 0.10? Tuesday 2.48pm 0.50? Positive breath alcohol results (mg/1000ml) 20

21 3.2.3 Breathalyser results for mine G1 Of the 111 samples obtained at this mine, 3 samples (2.7%) contained alcohol and one of these (0.9%) contained alcohol above the 0.10mg/1000ml limit as described in table 3.4. Table 3.4: Positive breathalyser results for mine G1 Day of sample collection Time of sample collection Positive breath alcohol results (mg/1000ml) Tuesday 10.28pm 0.08? Thursday 6.30am 0.13? Thursday 10.30am 0.07? Breathalysers results for mine G2 Of the 359 breath samples obtained, 12 samples (3.3%) contained alcohol, and the alcohol level in 7 of these samples (1.9%) was above the driving limit of 0.10mg/1000ml as depicted in table 3.5. Table 3.5: Positive breathalyser results for mine G2 Day of sample collection Time of sample collection Positive breath alcohol results (mg/1000ml) Monday 7:46pm 0.09? Monday 8:27pm 0.15? Monday 8:41pm 0.05? Tuesday 7:20pm 0.05? Wednesday 4:29am 0.27? Wednesday 4:58am 0.08? Wednesday 12:16pm 0.13? Wednesday 12.20pm 0.12? Wednesday 7.32pm 0.24? Wednesday 8.11pm 0.25? Friday 4.16am 0.05? Friday 6.52am 0.14? 21

22 3.2.5 Breathalyser results for mine D1 318 breath samples were obtained at this mine. Of the 5 samples (1.6%) that contained alcohol, 3 samples (0.9%) contained alcohol above the 0.10mg/1000ml driving limit, as described in table 3.6. Table 3.6: Positive breathalyser results for mine D1 Day of sample collection Time of sample collection Wednesday 6.53am 0.04? Wednesday 7.07am 0.24? Thursday 6.46am 0.04? Thursday 6.52am 0.22? Thursday 8.00am 0.16? Positive breath alcohol results (mg/1000ml) Breathalyser results for mine C1 266 breath samples were obtained at this mine. As described in table 3.7, 6 of these samples (2.3%) contained alcohol. 3 samples (1.1%) contained alcohol at or above the legal driving limit of 0.10mg/1000ml. Table 3.7: Positive breathalyser results for mine C1 Day of sample collection Time of sample collection Positive breath alcohol results (mg/1000ml) Tuesday 12.14pm 0.06? Wednesday 6.26am 0.07? Wednesday 10.13am 0.07? Wednesday 10.55am 0.22? Friday 9.45am 0.10? Friday 9.56am 0.11? 22

23 3.2.7 Breathalyser results for mine O1 204 breath samples were obtained from this mine. 16 of these samples (7.8%) contained alcohol, with 12 samples (5.9%) containing alcohol at or above the 0.10mg/1000ml legal limit as described in table 3.8, five of these 12 samples contained alcohol at more than four times the 0.10mg/1000ml limit with the highest reading of 0.88mg/1000ml being almost nine times the limit. Table 3.8: Positive breathalyser results for mine O1 Day of sample collection Time of sample collection Positive breath alcohol results (mg/1000ml) Monday 7.45am 0.48? Monday 7.51am 0.37? Monday 7.54am 0.11? Monday 7.58am 0.43? Monday 8.27am 0.21? Monday 9.07am 0.20? Monday 9.10am 0.10? Monday 9.18am 0.09? Monday 9.49am 0.06? Monday 9.51am 0.44? Monday 9.53am 0.03? Monday 10.51am 0.55? Monday 1.24pm 0.04? Monday 1.25pm 0.24? Monday 1.55pm 0.88? Monday 3.50pm 0.03? 23

24 3.2.8 Summary of breathalyser results of all study mines Overall, while 1.1% of samples obtained from all study mines contained alcohol below the 0.10mg/1000ml legal driving limit, 1.9% contained alcohol equal to or above this limit. 0% to 7.9% of all samples contained alcohol above the stipulated mine alcohol limit. According to the mine limit, all samples containing alcohol in mines C1 (2.2%) and O1 (7.9%) would be regarded as failed tests. Table 3.9 categorises the alcohol content of samples obtained from all mines with respect to the 0.10mg/1000ml legal driving limit and the mine alcohol limit. Table 3.9: Summary of breathalyser results by mine Mine Number of samples obtained % of samples with alcohol below 0.10mg/1000 ml legal driving limit % of samples with alcohol equal to or above 0.10mg/1000 ml legal driving limit Individual mine alcohol limits (mg/1000ml) % of sample with alcoho equal to or above the mine alcoho limit P % 0% (Nil) 0.2 0% (Nil) P % 1.5% (3) % (2) G % (2) 0.9% % (Nil) G % (5) 1.9% (7) % (3) D % (2) 0.9% (3) % C % (3) 1.1% (3) % (6) O % (4) 5.9% (12) % (16) Mean % (18) 1.9% (29) 1.8% (28) 24

25 3.3 Results of urine cannabis testing 9.1% of urine samples collected from all study mines tested positive for cannabis. Table 3.10 displays a summary of all positive results by mine. Table 3.10: Summary of urine test results by mine Mine Number of urine samples collected % of samples positive for cannabis P % (5) P % (15) G % (15) G % (20) D % (22) C % (20) O % (43) Mean % (140) Table 3.11 shows the proportion of cannabis positive urine samples in mines in the same commodity group. Table 3.11: Comparison of proportion of cannabis positive urine samples between commodity mines. Commodity Mine Proportion of cannabis positive urine samples P (n=307) 6.5% (20) G (n=468) 7.5% (35) D (n=305) 7.2% (22) C (n=262) 7.6% (20) O (n=200) 21.5% (43) 25